Abstract
BACKGROUND: Bipolar disorder is a recurrent and disabling condition, with a critical clinical need to prevent transitions from euthymia or depression (normal or low activation states) to mania (a high activation state). This study investigates how disruptions in sleep-wake and circadian rhythms may trigger these high activation states, to inform more effective relapse prevention strategies. METHODS: We developed a computational agent-based model integrating empirical evidence, clinical expertise, and lived experience to simulate how 24-hour sleep-wake behaviors (SWBs) influence manic episodes. Individual characteristics were drawn from the Brain and Mind Youth Cohort (N = 2,330), and multiple scenarios were simulated to assess how SWB dynamics affect the emergence and course of mania. RESULTS: In the absence of all irregularities, no individuals experienced a manic episode. Removing behavioral feedback loops resulted in a substantial reduction in manic episodes and delayed onset. In contrast, eliminating light-dark entrainment slightly increased the frequency of manic episodes, suggesting that seasonal adaptation plays a stabilizing role. When examining components of SWB separately, removing sleep irregularities alone had only a modest effect on mania rates, whereas reducing activity irregularities led to the largest benefit: a significant drop in mean manic episodes, a delay in onset, and preventing mania in 65% of the simulated agent population. CONCLUSIONS: Our findings highlight the value of computational modeling for uncovering causal dynamics in mental health. These specific findings demonstrate how daily irregularities in sleep-wake behavior may be a necessary condition for mania. Targeting behavioral regularity may offer a powerful pathway for prevention and early intervention.